The basic command works on cygwin (which is what I
would expect since it acts like unix).  However, the
same command fails on win32 with a message about
not talking to the gnuplot for 5sec so that is where the
IPC::Run (or is it open3) is a problem.

Cheers,
Chris

On Thu, Jan 26, 2012 at 1:35 PM, Craig DeForest
<[email protected]> wrote:
> Hmmm,  The "no curve option found that matches 'terminal' error sounds like a 
> basic problem with the test suite
> (which needs to be updated).  For a basic functionality test, try:
>
>
> % perl -Mblib /usr/bin/perldl #(or whatever)
> perl> use PDL::Graphics::Gnuplot;
> perl> $w=gpwin();
> perl> $w->plot(xvals(10),xvals(10)**2);
>
> which should generate no errors and a simple plot on your machine's default 
> device.
>
>
>
> On Jan 26, 2012, at 11:08 AM, Chris Marshall wrote:
>
>> I also get the same error on ASPerl 5.12.3 with Vista:
>>
>> F:\chm\pdl\git\PDL-Graphics-Gnuplot>dmake test
>> C:\Users\chris.h.marshall\local\asperl\bin\perl.exe
>> "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib\lib',
>> 'blib\arch')" t/*.
>> t
>> t/manifest.t .. skipped: Author tests not required for installation
>> t/plot.t ...... 1/5 testfile: pdl_graphics_gnuplot_test_Vonlcev
>> WARNING: Hmmm,  gnuplot didn't respond for 5 seconds.  I was expecting to 
>> read
>>   a version number.  Ah, well, I'm returning the object anyway -- but don't
>>   be surprised if it doesn't work.
>>
>> #   Failed test 'basic plotting succeeded without error'
>> #   at t/plot.t line 36.
>> # plot() died with 'No curve option found that matches 'terminal'
>> # '
>>
>> #   Failed test 'basic plotting created a reasonably-sized file'
>> #   at t/plot.t line 43.
>> # resulting output file should be ascii 79x24, but only contains 0 bytes
>>
>> #   Failed test 'error detection works'
>> #   at t/plot.t line 55.
>> # plot() produced no error
>> # Testing PDL::Graphics::Gnuplot 0.11ced, Perl 5.012003,
>> C:\Users\CHRISH~1.MAR\local\asperl\bin\perl.exe
>> # Looks like you failed 3 tests of 5.
>> t/plot.t ...... Dubious, test returned 3 (wstat 768, 0x300)
>> Failed 3/5 subtests
>>
>> Test Summary Report
>> -------------------
>> t/plot.t    (Wstat: 768 Tests: 5 Failed: 3)
>>  Failed tests:  2, 4-5
>>  Non-zero exit status: 3
>> Files=2, Tests=5,  2 wallclock secs ( 0.09 usr +  0.02 sys =  0.11 CPU)
>> Result: FAIL
>> Failed 1/2 test programs. 3/5 subtests failed.
>> dmake.exe:  Error code 131, while making 'test_dynamic'
>>
>> On Thu, Jan 26, 2012 at 10:18 AM, Chris Marshall <[email protected]> 
>> wrote:
>>> For what it is worth, IPC:Run 0.90 fails to pass all
>>> tests on cygwin/Vista and the latest git clone from
>>> Craig's repo also fails to pass tests.  Maybe something
>>> is broken in the test suite?
>>>
>>> --Chris
>>>
>>>  make test
>>> PERL_DL_NONLAZY=1 /usr/bin/perl.exe "-MExtUtils::Command::MM" "-e"
>>> "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
>>> t/manifest.t .. skipped: Author tests not required for installation
>>> t/plot.t ...... 1/5 testfile: pdl_graphics_gnuplot_test_8pt8eAN
>>>
>>> #   Failed test 'basic plotting succeeded without error'
>>> #   at t/plot.t line 36.
>>> # plot() died with 'No curve option found that matches 'terminal'
>>> # '
>>>
>>> #   Failed test 'basic plotting created a reasonably-sized file'
>>> #   at t/plot.t line 43.
>>> # resulting output file should be ascii 79x24, but only contains 0 bytes
>>>
>>> #   Failed test 'error detection works'
>>> #   at t/plot.t line 55.
>>> # plot() produced no error
>>> # Testing PDL::Graphics::Gnuplot 0.11ced, Perl 5.010001, /usr/bin/perl
>>> # Looks like you failed 3 tests of 5.
>>> t/plot.t ...... Failed 3/5 subtests
>>>
>>> Test Summary Report
>>> -------------------
>>> t/plot.t    (Wstat: 0 Tests: 5 Failed: 3)
>>>  Failed tests:  2, 4-5
>>> Files=2, Tests=5,  4 wallclock secs ( 0.03 usr  0.11 sys +  0.45 cusr
>>> 1.05 csys =  1.65 CPU)
>>> Result: FAIL
>>> Failed 1/2 test programs. 3/5 subtests failed.
>>> make: *** [test_dynamic] Error 255
>>>
>>>
>>> On Thu, Jan 26, 2012 at 9:50 AM, Chris Marshall <[email protected]> 
>>> wrote:
>>>> As far as PDL goes, the real problem is that the "talk
>>>> to gnuplot" part doesn't work for win32---at least
>>>> reliably.  Once the module is working for unix, macosx,
>>>> and windows it would be a candidate for migration
>>>> into the PDL core distribution.  Of course, there will
>>>> need to be something like Alien::Gnuplot to handle
>>>> installing the gnuplot dependency to support the
>>>> module.  I don't know if that is already done or not....
>>>>
>>>> --Chris
>>>>
>>>> On Thu, Jan 26, 2012 at 12:51 AM, Craig DeForest
>>>> <[email protected]> wrote:
>>>>> Thanks for having a look at this, Cliff.
>>>>>
>>>>> Both Dima and I left the default terminal to be the same as the built-in 
>>>>> Gnuplot default (that you get by not specifying one to gnuplot).  On the 
>>>>> Mac, this defaults to "aqua"; on Fedora Linux it defaults to "X11cairo" 
>>>>> (IIRC).
>>>>
>>>> What is the default on win32?  I know it isn't X11.
>>>>
>>>>
>>>>> On Jan 25, 2012, at 10:27 PM, Clifford Sobchuk wrote:
>>>>>
>>>>>> From your response, I take it you are using cygwin and not strawberry or 
>>>>>> activestate perl. The IPC methods, I think I saw them in the .pm file 
>>>>>> near the bottom. I was trying to figure out exactly what was being 
>>>>>> called. I saw that gnuplot itself was being called which does work (not 
>>>>>> with pipes unless they changed it to be consistent with *nix platforms - 
>>>>>> previously it had to call pgnuplot for any piped actions), but I didn't 
>>>>>> see where the terminal type was set (the default that is, I expect we 
>>>>>> can put any terminal that is supported by gnuplot). I guess that during 
>>>>>> the tests files it would call the X11 term type? Does this mean that if 
>>>>>> I installed Xming that it would magically work - or at least pass the 
>>>>>> tests?
>>>>>>
>>>>>> I can give that a try and I'll check the differences between the IPC of 
>>>>>> the Graphics::GnuPlotIF and PDL:Graphics::Gnuplot to see if there is 
>>>>>> anything that I can naively spot as important.
>>>>>>
>>>>>> I tried to build the .tar.gz file from the git and it fails during make.
>>>>>>
>>>>>> drzowie-PDL-Graphics-Gnuplot-cbcb72d>perl Makefile.PL
>>>>>> Set up gcc environment - 4.5.2
>>>>>> Checking if your kit is complete...
>>>>>> Warning: the following files are missing in your kit:
>>>>>>        debian/source
>>>>>> Please inform the author.
>>>>>> Writing Makefile for PDL::Graphics::Gnuplot
>>>>>> Writing MYMETA.yml and MYMETA.json
>>>>>>
>>>>>> C:\Users\eclisob\Downloads\drzowie-PDL-Graphics-Gnuplot-cbcb72d\drzowie-PDL-Graphics-Gnuplot-cbcb72d>make
>>>>>> process_begin: CreateProcess(NULL, rem, ...) failed.
>>>>>> make (e=2): The system cannot find the file specified.
>>>>>> make: *** [blibdirs] Error 2
>>>>>>
>>>>>> Thanks,
>>>>>> CLIFF SOBCHUK
>>>>>> Core RF Engineering
>>>>>> Phone 613-667-1974   ecn: 8109-71974
>>>>>> mobile 403-819-9233
>>>>>> yahoo: sobchuk
>>>>>> www.ericsson.com
>>>>>>
>>>>>> "The author works for Telefonaktiebolaget L M Ericsson ("Ericsson"), who 
>>>>>> is solely responsible for this email and its contents. All inquiries 
>>>>>> regarding this email should be addressed to Ericsson. The web site for 
>>>>>> Ericsson is www.ericsson.com."
>>>>>>
>>>>>> This Communication is Confidential. We only send and receive email on 
>>>>>> the basis of the terms set out at www.ericsson.com/email_disclaimer
>>>>>>
>>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Craig DeForest [mailto:[email protected]]
>>>>>> Sent: Wednesday, January 25, 2012 9:15 PM
>>>>>> To: Dima Kogan
>>>>>> Cc: Craig DeForest; Clifford Sobchuk; [email protected]
>>>>>> Subject: Re: [Perldl] PDL::Graphics::Gnuplot on Win32 Vista.
>>>>>>
>>>>>>
>>>>>> On Jan 25, 2012, at 4:43 PM, Dima Kogan wrote:
>>>>>>
>>>>>>>> Clifford Sobchuk <[email protected]> wrote:
>>>>>>>> I tried installing the PDL::Graphics::Gnuplot on top of 2.4.9 and it
>>>>>>>> resulted in the following report from cpanm. Any pointers?
>>>>>>>
>>>>>>> This just means that it couldn't run the gnuplot executable. I haven't
>>>>>>> made any effort to make it work on Windows, so that's why you are
>>>>>>> seeing the error. There shouldn't be a lot involved, but somebody has 
>>>>>>> to do it.
>>>>>>>
>>>>>>> Also it is important to note that the code on CPAN is my latest
>>>>>>> branch. Craig has done lots of work on this module, and you should use 
>>>>>>> his latest:
>>>>>>>
>>>>>>> https://github.com/drzowie/PDL-Graphics-Gnuplot
>>>>>>>
>>>>>>> I don't know if Windows support was something he touched, but it's 
>>>>>>> worth a try.
>>>>>>> Craig, care to weigh in?
>>>>>>
>>>>>> Sure.  I've been working down the parse tree for Gnuplot.  The existing 
>>>>>> module works great for a wide variety of simple things.  There is crude 
>>>>>> FITS support.  It is possible to do everything that gnuplot can do 
>>>>>> (which is quite a lot), but  edgier cases still require messing with the 
>>>>>> syntax a bit (e.g. "xrange" works as advertised [you pass in a list ref 
>>>>>> containing <min>,<max>], unless you set "xmdate" to plot time values on 
>>>>>> the X axis, in which case you have to manually pass in a string that 
>>>>>> contains double quotes -- but that will be fixed).
>>>>>>
>>>>>> As for Windows support, I find that the "x11" device works fine with the 
>>>>>> only Windows system I use.  Someone else who knows (or cares) how to 
>>>>>> support the cra^H^H^Hstuff that comes out of Redmond can hack the 2-3 
>>>>>> methods that do the actual IPC to the gnuplot process.
>>>>>>
>>>>>> Here's an example plot I made recently for a paper.  It shows fields of 
>>>>>> view of several instruments on the STEREO-A spacecraft and locations of 
>>>>>> several CMEs that were tracked at the end of 2008.  It was surprisingly 
>>>>>> easy to generate (and is just one frame in a movie).
>>>>>>
>>>>>> I'd like to get one or two more rounds of smartening finished up and a 
>>>>>> better test suite written, but I (and I think Dima) am planning to have 
>>>>>> a CPAN release of PDL::Graphics::Gnuplot buttoned up in time for folks 
>>>>>> to poke and prod it before 2.4.11 comes out (hopefully ~4-6 months after 
>>>>>> 2.4.10).
>>>>>>
>>>>>> Cheers,
>>>>>> Craig
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Perldl mailing list
>>>>> [email protected]
>>>>> http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
>>
>

_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to